An Intelligent Surveillance System Design Based On Embedded Technology

With the development of industrial process modernization,industrial enterprises have increasingly higher security requirements for the factory area.In recent years,with the continuous development of computer vision,the intelligent monitoring of the factory area has received more and more attention.At present,the processing core of intelligent monitoring system is mainly computer,which has the problems of poor integrity,high cost,large volume and difficult installation.Aiming at the existing problems,this paper carries out an in-depth study on the intelligent monitoring system based on factory embedded system.Firstly,the system chooses raspberry PI as the core processor of intelligent monitoring.Linux system and Open CV algorithm library are transplanted into the embedded development board.And the system realizes video acquisition,image processing,system alarm,video remote transmission and other basic functions.The whole system is not only low cost,but also has a wide range of application prospects.Secondly,this paper analyzes the safety problems existing in the factory and the current safety requirements,and realizes three kinds of intelligent detection functions.There are some dangerous areas in the factory,and according to the factory regulations,nopedestrians can appare in these places.The system realizes the pedestrian intrusion detection function for these areas.In some fire-prone areas,the system implements the function of flame detection.Some specific areas of the factory area require identity authentication before entering.For these areas,the system realizes the function of face recognition.This paper introduces in detail the implementation of three kinds of intelligent detection functions.Consideringthe processing performance of raspberry PI,this paper chooses the detection algorithm with low complexity,and performs simulation experiments on the algorithm to verify the accuracy and real-time performance of the system detection,which has high practical value.Finnally,the intelligent monitoring system designed in this paper takes embedded as the processor,and transmits the processed video data to the upper computer through the network,which makes the original secure video data easy to be tampered with or attacked by a third party.The monitoring video contains many images related to plant equipment and production.The intrusion of the video monitoring network will have serious consequences for the whole enterprise,such as leakage of production process and theft of commercial information.This paper analyzes the network traffic information in the transmission process.A network intrusion detection method combining information gain and extreme learning machine is proposed.This method uses information gain to reduce the dimension of network traffic characteristics.In order to make ELM better recognize network abnormal behavior,the system uses particle swarm optimization to optimize the parameters.Through the comparison experiment,it is concluded that the proposed method solves the network security problem of monitoring system from the perspective of practical application and has a high accuracy rate in identifying network intrusion.